AI Is Coming for America’s Cities. We Need a Plan.

Artificial intelligence will not reshape America evenly. It will hit hardest and fastest in our cities. 

This has been true of every major technological revolution. Steam, electricity, and the telegraph built the modern American city. The next wave—containerized shipping, interstate highways, horizontal factories—led to urban decline. The information revolution that followed brought a renaissance to some cities and left others further behind.

Now AI will reshape cities again. It will land first in the office towers full of lawyers, accountants, consultants, analysts, and designers who power the modern urban economy. If AI augments their work, cities will grow stronger. If it instead replaces them, there will be painful job loss, and the office economy that sustains city budgets will come under serious strain.

The same is true of the institutions and infrastructure of cities. Large urban hospitals could use AI to improve diagnosis, discovery, and care—though the gains may not reach every patient or every worker equally. Big-city school systems could use it to offer every child personalized instruction, or see it widen the learning gaps they are already struggling to close.

No city is more exposed to the promise and peril of AI than New York. We have more white-collar workers than anywhere in America, and they are concentrated in the sectors most exposed to disruption. New York’s investment banks and private equity firms have financed the AI boom and would feel the shock if it falters. Our City budget depends heavily on tax revenue generated by this new economy, bringing significant upside but also real risk. And in a city marked by some of the deepest inequality in the nation, AI could widen the wealth divide even further.

We are already starting to feel some of these effects. Job gains in New York City have been sluggish, even amid a booming stock market. The unemployment rate for New Yorkers in their 20s is now higher for those with a college degree than those without one, a reversal of a pattern going back generations.  And the search for entry-level, white-collar jobs here has become simply brutal (as any young New Yorker will tell you).

In a better world the federal government would be moving with urgency to help cities, and the nation as a whole, manage this transition. It would set safety standards for advanced AI systems, plan for displaced workers, tax the extraordinary wealth generated by AI so gains are shared broadly and ensure democratic control of this world-changing technology.

Alas, the Trump administration has made clear that federal oversight of AI will be limited at best, and Congress has done virtually nothing to fill the gap. So it will fall on local government to act. Fortunately, we have many tools at our disposal.

The first step is to get a better fix on what’s coming. Last month, my office published projections of AI’s impact over the next 3-5 years on New York City’s economy, job market, and budget. Because the pace and impact of AI remain deeply uncertain, the analysis presented five planning scenarios with dramatically different implications for New York City.

In the more positive scenarios, AI drives productivity gains and financial market growth while creating more jobs than it eliminates—a net yearly gain of nearly 58,000 private sector jobs citywide in the most optimistic case. But net growth does not mean no one is hurt. Workers displaced from legacy roles often cannot easily move into the new jobs being created.

In the more negative scenarios, AI replaces jobs faster than new ones are created. Our most dire forecast projects a rapid loss of 250,000 jobs in New York City. We also forecast a scenario in which the AI investment boom bursts, leading markets to retreat, and precipitating a recession. Under these scenarios New York City’s budget will take a severe hit, with tax revenues below their growth path by as much as $14.4 billion.

Overall, the scenarios have an even 50/50 probability of more positive vs. negative economic and fiscal impacts on New York. So what is a city to do with such an uncertain AI future?

For cities exposed to AI-driven disruption, the most immediate task is to build a fiscal buffer for whatever lies ahead. New York City currently trails every major American city except Chicago in reserves relative to total revenues. Stronger reserves would let us protect essential services, support workers displaced by AI, and respond creatively to the turbulence ahead.

We also need to prepare City government itself for the challenges ahead. New York’s vast municipal machinery rests on a rickety foundation of 1980s and 1990s technology—systems wholly unprepared for the challenges and opportunities AI will bring. We need to modernize those systems, so City services are faster, more efficient, and more humane. We need stronger defenses against AI-enabled cyberattacks, which are already growing noticeably more sophisticated.

New York City’s $320 billion pension fund makes us one of the largest shareholders in corporate America, and we should use that power. Through shareholder resolutions, proxy votes, and other forms of pressure, we should demand that companies deploy and govern AI responsibly.

And we should grow New York City’s already strong ecosystem of companies building AI applications. That is how we share in the upside as the sector expands, and how we shape the technology ourselves rather than just live with what others build.

There will be resistance to every one of these strategies, in New York and in cities around the nation. The backlash against AI is now so intense—not without justification—that even engaging with the question of how to manage this juggernaut has become politically difficult.

But surrendering decisions about our future to Silicon Valley, Washington, or the market alone would be a grave mistake. 

This is a lesson cities should have learned by now. Those that ignored containerization in the 1960s lost their ports. Those that embraced the information revolution in the 2000s became global hubs. Now, facing the most consequential technology in history, cities must choose action.